{"id":190180,"date":"2023-02-13T10:41:39","date_gmt":"2023-02-13T10:41:39","guid":{"rendered":"https:\/\/datascientest.com\/?p=190180"},"modified":"2023-11-09T13:54:07","modified_gmt":"2023-11-09T13:54:07","slug":"fastai-tout-savoir","status":"publish","type":"post","link":"https:\/\/datascientest.com\/fastai-tout-savoir","title":{"rendered":"FastAI : Une API de Deep Learning accessible \u00e0 tous"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-post\" data-elementor-id=\"190180\" class=\"elementor elementor-190180\" data-elementor-post-type=\"post\">\n\t\t\t\t\t\t\t<div class=\"elementor-element elementor-element-5e6d7d9 e-flex e-con-boxed e-con\" data-id=\"5e6d7d9\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-4a2252b elementor-widget elementor-widget-heading\" data-id=\"4a2252b\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-heading-title{padding:0;margin:0;line-height:1}.elementor-widget-heading .elementor-heading-title[class*=elementor-size-]>a{color:inherit;font-size:inherit;line-height:inherit}.elementor-widget-heading .elementor-heading-title.elementor-size-small{font-size:15px}.elementor-widget-heading .elementor-heading-title.elementor-size-medium{font-size:19px}.elementor-widget-heading .elementor-heading-title.elementor-size-large{font-size:29px}.elementor-widget-heading .elementor-heading-title.elementor-size-xl{font-size:39px}.elementor-widget-heading .elementor-heading-title.elementor-size-xxl{font-size:59px}<\/style><h2 class=\"elementor-heading-title elementor-size-default\">FastAI est une biblioth\u00e8que open source tr\u00e8s intuitive comprenant un vaste panel de fonctions utiles \u00e0 un projet de Deep Learning. Cet article pr\u00e9sente une vue d\u2019ensemble des modules disponibles ainsi que les avantages que proposent FastAI.<\/h2>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-9f983b3 e-flex e-con-boxed e-con\" data-id=\"9f983b3\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-7d096c6 elementor-widget elementor-widget-heading\" data-id=\"7d096c6\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Qu\u2019est-ce que FastAI ?<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-75efc4e elementor-widget elementor-widget-text-editor\" data-id=\"75efc4e\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-text-editor.elementor-drop-cap-view-stacked .elementor-drop-cap{background-color:#69727d;color:#fff}.elementor-widget-text-editor.elementor-drop-cap-view-framed .elementor-drop-cap{color:#69727d;border:3px solid;background-color:transparent}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap{margin-top:8px}.elementor-widget-text-editor:not(.elementor-drop-cap-view-default) .elementor-drop-cap-letter{width:1em;height:1em}.elementor-widget-text-editor .elementor-drop-cap{float:left;text-align:center;line-height:1;font-size:50px}.elementor-widget-text-editor .elementor-drop-cap-letter{display:inline-block}<\/style>\t\t\t\t<p>D\u00e9ploy\u00e9e en 2018, <strong>FastAI<\/strong> est une librairie Python de Deep Learning construite \u00e0 partir de <a href=\"https:\/\/datascientest.com\/pytorch-tout-savoir\">PyTorch<\/a>. Il s\u2019agit d\u2019une librairie open source, dont le repository est disponible sur <a href=\"https:\/\/github.com\/fastai\/fastai\">GitHub<\/a>.<\/p><p>Le groupe fast.ai poss\u00e8de par ailleurs son propre site, <strong>aliment\u00e9 d\u2019un<\/strong> <a href=\"https:\/\/datascientest.com\/quest-ce-quun-mooc\">MOOC<\/a> et de la documentation du package.<\/p><p><span style=\"font-weight: 400;\">L\u2019attrait principal de FastAI r\u00e9side dans ses nombreux composants haut-niveau int\u00e9gr\u00e9s, <strong>permettant un d\u00e9ploiement rapide des architectures Deep Learning<\/strong>. Il est trivial de reproduire des r\u00e9sultats fondamentaux de Deep Learning, le tout avec tr\u00e8s peu de lignes.<\/span><\/p><p><span style=\"font-weight: 400;\">\u00c0 contrario, il existe pl\u00e9thore de fonctions bas-niveau, donnant la possibilit\u00e9 de manipuler, cr\u00e9er \u00e0 sa guise de nouveaux mod\u00e8les personnalis\u00e9s.<\/span><\/p><p><span style=\"font-weight: 400;\">FastAI regroupe <strong>la clart\u00e9 et la rapidit\u00e9 d\u2019ex\u00e9cution de <\/strong><a href=\"https:\/\/datascientest.com\/keras\">Keras<\/a> <strong>avec la flexibilit\u00e9 de PyTorch<\/strong>, formant un produit \u00e9pur\u00e9 et maniable. C\u2019est avec cet \u00e9tat d\u2019esprit que la biblioth\u00e8que est structur\u00e9e, comme nous le verrons dans la section suivante.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-36fd37b e-flex e-con-boxed e-con\" data-id=\"36fd37b\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-82eeadb elementor-widget elementor-widget-heading\" data-id=\"82eeadb\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Une API unique<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-bb5ca7e elementor-widget elementor-widget-text-editor\" data-id=\"bb5ca7e\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Un aspect important de FastAI est la <\/span><b>centralisation des modules<\/b><span style=\"font-weight: 400;\">. En g\u00e9n\u00e9ral, il est de bon usage de n\u2019importer que les fonctions n\u00e9cessaires au projet ; cela peut requ\u00e9rir un effort suppl\u00e9mentaire qui peut tr\u00e8s vite s\u2019accro\u00eetre \u00e0 mesure que le projet avance.<\/span><\/p><p><b>FastAI <\/b><span style=\"font-weight: 400;\">\u00e9vince cette contrainte : il suffit d\u2019importer l\u2019int\u00e9gralit\u00e9 du ou des modules qui nous concernent. On verra souvent les importations suivantes qui forment le coeur de l\u2019API :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4b4c46e elementor-widget elementor-widget-shortcode\" data-id=\"4b4c46e\" data-element_type=\"widget\" data-widget_type=\"shortcode.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-shortcode\"><script src=\"https:\/\/gist.github.com\/victorDST\/31420946d9cd1056d5b597e525f2884a.js\"><\/script><\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-4881a2d elementor-widget elementor-widget-text-editor\" data-id=\"4881a2d\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Plus besoin alors de sp\u00e9cifier les fonctions \u00e0 importer, contrairement \u00e0 keras dont les performances vont de pair avec la ma\u00eetrise de l\u2019utilisateur.\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">En pratique, FastAI agit comme une API centralis\u00e9e, regroupant toutes sortes de fonctions utiles \u00e0 un projet de Deep Learning. A haut niveau, il est possible de travailler sur quatre domaines fondamentaux : Computer Vision, NLP, filtrage collaboratif et tabulation de donn\u00e9es (analyse temporelle).<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-2a8765c elementor-widget elementor-widget-image\" data-id=\"2a8765c\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<style>\/*! elementor - v3.15.0 - 02-08-2023 *\/\n.elementor-widget-image{text-align:center}.elementor-widget-image a{display:inline-block}.elementor-widget-image a img[src$=\".svg\"]{width:48px}.elementor-widget-image img{vertical-align:middle;display:inline-block}<\/style>\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"683\" height=\"773\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/Fichier-100.png\" class=\"attachment-full size-full wp-image-190210\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/Fichier-100.png 683w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/Fichier-100-265x300.png 265w\" sizes=\"(max-width: 683px) 100vw, 683px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e0500b5 elementor-widget elementor-widget-text-editor\" data-id=\"e0500b5\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Le diagramme ci-dessus pr\u00e9sente l\u2019architecture hi\u00e9rarchis\u00e9e de FastAI. Si l\u2019utilisateur souhaite modifier une partie de l\u2019API haut niveau, il n\u2019a pas besoin de conna\u00eetre en amont le fonctionnement au plus bas niveau du mod\u00e8le.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-0580e4a e-flex e-con-boxed e-con\" data-id=\"0580e4a\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-1df2b48 elementor-widget elementor-widget-heading\" data-id=\"1df2b48\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Une d\u00e9mocratisation du Deep Learning<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-9171a3c elementor-widget elementor-widget-text-editor\" data-id=\"9171a3c\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Selon <strong>Jeremy Howard<\/strong> et<strong> Rachel Thomas<\/strong> (fondateurs de fast.ai), l\u2019objectif derri\u00e8re cette uniformisation est de d\u00e9mocratiser le Deep Learning afin de faciliter son acc\u00e8s au grand public. Pour respecter cette mission, la syntaxe est sensiblement moins prolixe que les autres biblioth\u00e8ques de r\u00e9f\u00e9rence : quelques lignes seulement sont n\u00e9cessaires pour cr\u00e9er un mod\u00e8le de z\u00e9ro.<\/span><\/p><p><span style=\"font-weight: 400;\">Parmi les efforts fournis, il existe notamment une unique classe <\/span><b>Learner<\/b><span style=\"font-weight: 400;\"> qui prend en compte l\u2019architecture du mod\u00e8le, la data et l\u2019optimiseur. FastAI propose des valeurs par d\u00e9faut pertinentes, et choisit \u00e0 notre place la fonction de co\u00fbt adapt\u00e9e aux besoins. Citons \u00e9galement la classe <\/span><b>DataLoader <\/b><span style=\"font-weight: 400;\">qui construit automatiquement les sets de validation\/test, induisant les \u00e9tapes de <strong>preprocessing<\/strong> (m\u00e9lange des donn\u00e9es, stratification). Une aubaine pour un d\u00e9butant qui oublierait de mettre de c\u00f4t\u00e9 une partie de son jeu de donn\u00e9es\u2026<\/span><\/p><p><span style=\"font-weight: 400;\">Comme mentionn\u00e9 pr\u00e9c\u00e9demment, il est tout \u00e0 fait possible pour les plus aguerris d\u2019acc\u00e9der aux couches inf\u00e9rieures de l\u2019<a href=\"https:\/\/datascientest.com\/api\">API<\/a> et de le modifier \u00e0 sa guise sans souci.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-2bad274 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"2bad274\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formation-deep-learning\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">En savoir plus sur la formation au Deep Learning<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-c414c2d e-flex e-con-boxed e-con\" data-id=\"c414c2d\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-9d407c1 elementor-widget elementor-widget-heading\" data-id=\"9d407c1\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Un exemple d\u2019application en Computer Vision<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-90f53a2 elementor-widget elementor-widget-text-editor\" data-id=\"90f53a2\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Prenons l\u2019exemple d\u2019un r\u00e9seau de neurones convolutifs (ou Convolutional Neural Network, abr\u00e9g\u00e9 <\/span><strong>CNN<\/strong><span style=\"font-weight: 400;\">). Ce genre de mod\u00e8le est couramment utilis\u00e9 en traitement d\u2019images, comme par exemple en reconnaissance de features.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-af82367 elementor-widget elementor-widget-image\" data-id=\"af82367\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"1920\" height=\"540\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/convolutif-1.png\" class=\"attachment-full size-full wp-image-190211\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/convolutif-1.png 1920w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/convolutif-1-300x84.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/convolutif-1-1024x288.png 1024w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/convolutif-1-768x216.png 768w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/convolutif-1-1536x432.png 1536w\" sizes=\"(max-width: 1920px) 100vw, 1920px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-167b418 elementor-widget elementor-widget-text-editor\" data-id=\"167b418\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Pour en savoir plus sur le fonctionnement des CNN, vous pouvez consulter <a href=\"https:\/\/datascientest.com\/convolutional-neural-network\">notre dossier d\u00e9di\u00e9<\/a>.<\/span><\/p><p><span style=\"font-weight: 400;\">Avec FastAI, voici tout ce qu\u2019il faut \u00e9crire pour r\u00e9aliser notre mod\u00e8le :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-db4581a elementor-widget elementor-widget-shortcode\" data-id=\"db4581a\" data-element_type=\"widget\" data-widget_type=\"shortcode.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-shortcode\"><script src=\"https:\/\/gist.github.com\/victorDST\/bf3efa2a36cd000503eac5608caa8601.js\"><\/script><\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-52df617 elementor-widget elementor-widget-text-editor\" data-id=\"52df617\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Comme vous pouvez le remarquer, certains param\u00e8tres ne sont pas sp\u00e9cifi\u00e9s ; <\/span><i><span style=\"font-weight: 400;\">learn <\/span><\/i><span style=\"font-weight: 400;\">s\u2019occupe d\u2019attribuer les bonnes valeurs \u00e0 ceux-ci. Il va t\u00e9l\u00e9charger un mod\u00e8le pr\u00e9-entra\u00een\u00e9 sur ImageNet, et initialiser entre autres l\u2019optimiseur, le taux d\u2019apprentissage, le weight decay, etc., sauf si l\u2019utilisateur d\u00e9cide de renseigner ces donn\u00e9es manuellement.<\/span><\/p><p><span style=\"font-weight: 400;\">Pour observer un \u00e9chantillon de notre jeu de donn\u00e9es, il suffit d\u2019appeler la fonction :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e2b6cad elementor-widget elementor-widget-shortcode\" data-id=\"e2b6cad\" data-element_type=\"widget\" data-widget_type=\"shortcode.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-shortcode\"><script src=\"https:\/\/gist.github.com\/victorDST\/74e19a07ccadef2d1a740d735ef731e2.js\"><\/script><\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-f8fcf5b elementor-widget elementor-widget-image\" data-id=\"f8fcf5b\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"522\" height=\"523\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image5-1.png\" class=\"attachment-full size-full wp-image-190194\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image5-1.png 522w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image5-1-300x300.png 300w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image5-1-150x150.png 150w\" sizes=\"(max-width: 522px) 100vw, 522px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-dca7850 elementor-widget elementor-widget-text-editor\" data-id=\"dca7850\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\"><strong>show_batch()<\/strong> fournit une visualisation de la dara diff\u00e9rente selon la forme de cette derni\u00e8re.<\/span><\/p><p><span style=\"font-weight: 400;\">Pour entra\u00eener le mod\u00e8le et v\u00e9rifier ses pr\u00e9dictions, deux lignes viennent compl\u00e9ter notre notebook :<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-1602319 elementor-widget elementor-widget-shortcode\" data-id=\"1602319\" data-element_type=\"widget\" data-widget_type=\"shortcode.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-shortcode\"><script src=\"https:\/\/gist.github.com\/victorDST\/8a035098a11a7234779c0fe0bd364712.js\"><\/script><\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-5741e68 elementor-widget elementor-widget-image\" data-id=\"5741e68\" data-element_type=\"widget\" data-widget_type=\"image.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<img decoding=\"async\" width=\"516\" height=\"193\" src=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image4-1.png\" class=\"attachment-full size-full wp-image-190198\" alt=\"\" loading=\"lazy\" srcset=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image4-1.png 516w, https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/image4-1-300x112.png 300w\" sizes=\"(max-width: 516px) 100vw, 516px\" \/>\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-e92fa95 elementor-widget elementor-widget-text-editor\" data-id=\"e92fa95\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Avec tr\u00e8s peu de lignes de code, nous sommes parvenus \u00e0 cr\u00e9er un CNN op\u00e9rationnel. Bluffant, n\u2019est-ce pas ?\u00a0<\/span><\/p><p><span style=\"font-weight: 400;\">Pour aller plus loin, sachez qu\u2019il existe un syst\u00e8me tr\u00e8s intuitif et modulable de 2-way <\/span><b>callback<\/b><span style=\"font-weight: 400;\">,<\/span> <span style=\"font-weight: 400;\">permettant de contr\u00f4ler le d\u00e9roulement de l\u2019entra\u00eenement \u00e0 souhait.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-e46630d e-flex e-con-boxed e-con\" data-id=\"e46630d\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-58b6d99 elementor-widget elementor-widget-heading\" data-id=\"58b6d99\" data-element_type=\"widget\" data-widget_type=\"heading.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t<h3 class=\"elementor-heading-title elementor-size-default\">Le mot de la fin<\/h3>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-cdd7694 elementor-widget elementor-widget-text-editor\" data-id=\"cdd7694\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">En r\u00e9sum\u00e9, <\/span><b>FastAI <\/b><span style=\"font-weight: 400;\">adopte une approche \u00e9pur\u00e9e pour g\u00e9rer des projets de Deep Learning, gr\u00e2ce \u00e0 une biblioth\u00e8que riche en fonctions haut-niveau. Tr\u00e8s facile \u00e0 prendre en main, l\u2019API est disponible \u00e0 l\u2019importation sur Jupyter. N\u2019h\u00e9sitez pas \u00e0 tester par vous-m\u00eame, et \u00e0 vous r\u00e9f\u00e9rer si besoin \u00e0 <a href=\"https:\/\/www.fast.ai\/posts\/2020-02-13-fastai-A-Layered-API-for-Deep-Learning.html\">la documentation du site officiel<\/a>.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t<div class=\"elementor-element elementor-element-54bfe75 e-flex e-con-boxed e-con\" data-id=\"54bfe75\" data-element_type=\"container\" data-settings=\"{&quot;content_width&quot;:&quot;boxed&quot;}\">\n\t\t\t\t\t<div class=\"e-con-inner\">\n\t\t\t\t<div class=\"elementor-element elementor-element-30a20d2 elementor-widget elementor-widget-text-editor\" data-id=\"30a20d2\" data-element_type=\"widget\" data-widget_type=\"text-editor.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t\t\t<p><span style=\"font-weight: 400;\">Nous esp\u00e9rons que cet article vous a plu, et si vous souhaitez aller plus loin en Deep Learning, d\u00e9couvrez sans plus tarder nos cursus experts : <a href=\"https:\/\/datascientest.com\/formation-deep-learning\/formation-computer-vision\">Computer Vision<\/a> et <a href=\"https:\/\/datascientest.com\/formation-deep-learning\/natural-language-processing\">Natural Language Processing<\/a>.<\/span><\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"elementor-element elementor-element-ad98f87 elementor-align-center button-principale elementor-widget elementor-widget-button\" data-id=\"ad98f87\" data-element_type=\"widget\" data-widget_type=\"button.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<div class=\"elementor-button-wrapper\">\n\t\t\t<a class=\"elementor-button elementor-button-link elementor-size-sm elementor-animation-grow\" href=\"https:\/\/datascientest.com\/formation-deep-learning\">\n\t\t\t\t\t\t<span class=\"elementor-button-content-wrapper\">\n\t\t\t\t\t\t<span class=\"elementor-button-text\">Je me forme au Deep Learning<\/span>\n\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>FastAI est une biblioth\u00e8que open source tr\u00e8s intuitive comprenant un vaste panel de fonctions utiles \u00e0 un projet de Deep Learning. Cet article pr\u00e9sente une vue d\u2019ensemble des modules disponibles ainsi que les avantages que proposent FastAI. Qu\u2019est-ce que FastAI ? D\u00e9ploy\u00e9e en 2018, FastAI est une librairie Python de Deep Learning construite \u00e0 partir [&hellip;]<\/p>\n","protected":false},"author":78,"featured_media":190209,"comment_status":"open","ping_status":"open","sticky":false,"template":"elementor_theme","format":"standard","meta":{"content-type":"","inline_featured_image":false},"categories":[2497],"tags":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v21.7 (Yoast SEO v21.7) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>FastAI : Une API de Deep Learning accessible \u00e0 tous | DataScientest.com<\/title>\n<meta name=\"description\" content=\"FastAI est une biblioth\u00e8que open source tr\u00e8s intuitive comprenant un vaste panel de fonctions utiles \u00e0 un projet de Deep Learning. Cet article pr\u00e9sente une vue d\u2019ensemble des modules disponibles ainsi que les avantages que proposent FastAI.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/datascientest.com\/fastai-tout-savoir\" \/>\n<meta property=\"og:locale\" content=\"fr_FR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"FastAI : Une API de Deep Learning accessible \u00e0 tous\" \/>\n<meta property=\"og:url\" content=\"https:\/\/datascientest.com\/fastai-tout-savoir\" \/>\n<meta property=\"og:site_name\" content=\"Formation Data Science | DataScientest.com\" \/>\n<meta property=\"article:published_time\" content=\"2023-02-13T10:41:39+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-11-09T13:54:07+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/Illu_BLOG_fastAI.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1024\" \/>\n\t<meta property=\"og:image:height\" content=\"563\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"J\u00e9r\u00e9my Robert\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u00c9crit par\" \/>\n\t<meta name=\"twitter:data1\" content=\"J\u00e9r\u00e9my Robert\" \/>\n\t<meta name=\"twitter:label2\" content=\"Dur\u00e9e de lecture estim\u00e9e\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/datascientest.com\/fastai-tout-savoir\",\"url\":\"https:\/\/datascientest.com\/fastai-tout-savoir\",\"name\":\"FastAI : Une API de Deep Learning accessible \u00e0 tous | DataScientest.com\",\"isPartOf\":{\"@id\":\"https:\/\/datascientest.com\/#website\"},\"datePublished\":\"2023-02-13T10:41:39+00:00\",\"dateModified\":\"2023-11-09T13:54:07+00:00\",\"author\":{\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925\"},\"description\":\"FastAI est une biblioth\u00e8que open source tr\u00e8s intuitive comprenant un vaste panel de fonctions utiles \u00e0 un projet de Deep Learning. Cet article pr\u00e9sente une vue d\u2019ensemble des modules disponibles ainsi que les avantages que proposent FastAI.\",\"inLanguage\":\"fr-FR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/datascientest.com\/fastai-tout-savoir\"]}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/datascientest.com\/#website\",\"url\":\"https:\/\/datascientest.com\/\",\"name\":\"Formation Data Science | DataScientest.com\",\"description\":\"Leader de la formation en Data Science\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/datascientest.com\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"fr-FR\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925\",\"name\":\"J\u00e9r\u00e9my Robert\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"fr-FR\",\"@id\":\"https:\/\/datascientest.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg\",\"contentUrl\":\"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg\",\"caption\":\"J\u00e9r\u00e9my Robert\"},\"description\":\"J\u00e9r\u00e9my Robert est actuellement Directeur p\u00e9dagogique B2B chez DataScientest.com. Avant de rejoindre DataScientest, il a travaill\u00e9 en tant que Consultant Business Intelligence chez DEAPDATA, o\u00f9 il a g\u00e9r\u00e9 des projets dans les domaines de l'assurance, bancaire et industrie. Il a \u00e9galement \u00e9t\u00e9 Co-fondateur et pr\u00e9sident de Tripzer. En tant que consultant ind\u00e9pendant, J\u00e9r\u00e9my a propos\u00e9 des formations sur MicroStrategy pour divers niveaux professionnels. Il est dipl\u00f4m\u00e9 de l'Universit\u00e9 Paris Cit\u00e9 avec un Master MIAGE, combinant comp\u00e9tences en informatique et gestion.\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/jeremyrobert1\/\",\"Data Science\"],\"url\":\"https:\/\/datascientest.com\/author\/jeremyrobert\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"FastAI : Une API de Deep Learning accessible \u00e0 tous | DataScientest.com","description":"FastAI est une biblioth\u00e8que open source tr\u00e8s intuitive comprenant un vaste panel de fonctions utiles \u00e0 un projet de Deep Learning. Cet article pr\u00e9sente une vue d\u2019ensemble des modules disponibles ainsi que les avantages que proposent FastAI.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/datascientest.com\/fastai-tout-savoir","og_locale":"fr_FR","og_type":"article","og_title":"FastAI : Une API de Deep Learning accessible \u00e0 tous","og_url":"https:\/\/datascientest.com\/fastai-tout-savoir","og_site_name":"Formation Data Science | DataScientest.com","article_published_time":"2023-02-13T10:41:39+00:00","article_modified_time":"2023-11-09T13:54:07+00:00","og_image":[{"width":1024,"height":563,"url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/02\/Illu_BLOG_fastAI.png","type":"image\/png"}],"author":"J\u00e9r\u00e9my Robert","twitter_card":"summary_large_image","twitter_misc":{"\u00c9crit par":"J\u00e9r\u00e9my Robert","Dur\u00e9e de lecture estim\u00e9e":"6 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/datascientest.com\/fastai-tout-savoir","url":"https:\/\/datascientest.com\/fastai-tout-savoir","name":"FastAI : Une API de Deep Learning accessible \u00e0 tous | DataScientest.com","isPartOf":{"@id":"https:\/\/datascientest.com\/#website"},"datePublished":"2023-02-13T10:41:39+00:00","dateModified":"2023-11-09T13:54:07+00:00","author":{"@id":"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925"},"description":"FastAI est une biblioth\u00e8que open source tr\u00e8s intuitive comprenant un vaste panel de fonctions utiles \u00e0 un projet de Deep Learning. Cet article pr\u00e9sente une vue d\u2019ensemble des modules disponibles ainsi que les avantages que proposent FastAI.","inLanguage":"fr-FR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/datascientest.com\/fastai-tout-savoir"]}]},{"@type":"WebSite","@id":"https:\/\/datascientest.com\/#website","url":"https:\/\/datascientest.com\/","name":"Formation Data Science | DataScientest.com","description":"Leader de la formation en Data Science","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/datascientest.com\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"fr-FR"},{"@type":"Person","@id":"https:\/\/datascientest.com\/#\/schema\/person\/7af2361c48e24e79a8aa2360e6e74925","name":"J\u00e9r\u00e9my Robert","image":{"@type":"ImageObject","inLanguage":"fr-FR","@id":"https:\/\/datascientest.com\/#\/schema\/person\/image\/","url":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg","contentUrl":"https:\/\/datascientest.com\/wp-content\/uploads\/2023\/11\/jeremyrobert.jpeg","caption":"J\u00e9r\u00e9my Robert"},"description":"J\u00e9r\u00e9my Robert est actuellement Directeur p\u00e9dagogique B2B chez DataScientest.com. Avant de rejoindre DataScientest, il a travaill\u00e9 en tant que Consultant Business Intelligence chez DEAPDATA, o\u00f9 il a g\u00e9r\u00e9 des projets dans les domaines de l'assurance, bancaire et industrie. Il a \u00e9galement \u00e9t\u00e9 Co-fondateur et pr\u00e9sident de Tripzer. En tant que consultant ind\u00e9pendant, J\u00e9r\u00e9my a propos\u00e9 des formations sur MicroStrategy pour divers niveaux professionnels. Il est dipl\u00f4m\u00e9 de l'Universit\u00e9 Paris Cit\u00e9 avec un Master MIAGE, combinant comp\u00e9tences en informatique et gestion.","sameAs":["https:\/\/www.linkedin.com\/in\/jeremyrobert1\/","Data Science"],"url":"https:\/\/datascientest.com\/author\/jeremyrobert"}]}},"_links":{"self":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/190180"}],"collection":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/users\/78"}],"replies":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/comments?post=190180"}],"version-history":[{"count":0,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/posts\/190180\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media\/190209"}],"wp:attachment":[{"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/media?parent=190180"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/categories?post=190180"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datascientest.com\/wp-json\/wp\/v2\/tags?post=190180"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}